This PDF file contains the front matter associated with SPIE Proceedings Volume 11218, including the Title Page, Copyright information, Table of Contents, Author and Conference Committee lists.

We proposed a novel oblique scanning laser ophthalmoscopy (oSLO) for the human eye. By using angled excitation light sheet and detection path, oSLO can dramatically improve the depth resolution and accelerate the imaging speed. With a novel optical design based on cylindrical lens, we break up the limitation raised by the small angle of the excitation light sheet owing to the small numerical aperture of the human eye. The current setup can image a human eye model over a field of view of 6 mm × 2.8 mm × 0.8mm with 7 µm transverse resolution and 41 µm axial resolution.

Optical coherence tomography (OCT) reveals the depth-resolved structure of the posterior eye non-invasively. However, artifacts caused by involuntary eye movement is one of the largest problems. Recently, we have demonstrated a motionartifact- free, high-resolution imaging technique based on Lissajous scanning pattern and advanced motion correction algorithm. Although this method works to a certain degree, the residual artifacts are still problematic for clinical applications.

In this study, we demonstrate the improvement of motion correction for en-face OCT imaging. The OCT signals are acquired with a Lissajous scanning pattern which has been modified from a standard Lissajous scan to enable OCT angiography (OCT-A) imaging. The lateral motion is estimated from several en-face images of OCT and OCT-A by using a motion estimation algorithm. Some diseased eyes exhibit abnormal patterns in OCT en-face images. Simultaneously using these images will enhance the motion estimation and will improve the motion correction at these abnormal regions. Motion-free imaging for retinal diseases is demonstrated.

Artificial optical stimulation of retinal neurons offers a potentially powerful approach towards the restoration of vision following photoreceptor loss in retinal degeneration. To better realize the potential of this approach, we introduced a holographic wavefront shaping method suitable for power-efficient patterned stimulation of retinal neurons. Here, to advance towards human translation of this technology, we develop and study a holographic interface with the human eye, designed to achieve cellular resolution stimulation in real-time. To this end, we first design and construct an optimized holographic display for high acuity optical retinal prosthesis. Next, we study different speckle-elimination strategies and adapt them to the projection of Sloan letters, a standard visual performance test. Finally, we perform psycho-physical experiments on normally sighted individuals to characterize and validate the performance of our display, and provide evidence for the ability of subjects aided by the display to perform high-acuity demanding visual tasks in versatile spatial and temporal illumination conditions.

Detecting and quantifying choroidal neovascularization (CNV) is essential for the diagnosis of neovascular age-related macular degeneration (AMD). Projection-resolved OCT angiography (PR-OCTA) has enabled both en face and volumetric visualization of CNV. However, previously described CNV detection methods only quantify CNV that was already diagnosed, and were unable to identify CNV form unknown inputs . Previous methods were also limited by artifacts. A fully automated CNV diagnosis and quantification algorithm using convolutional neural networks (CNNs) was developed. It was able to diagnose CNV and output CNV membrane and vessel area from retinal structural and angiographic images.

Although vis-OCT oximetry has been successfully demonstrated in vivo in human retina, it is so far limited on large vessels. To precisely locate the small vessels in depth dimension, OCTA is the natural method since it can effectively enhance the vessel contrast. In this paper, we demonstrate the first visible OCTA image in human retina, the small blood vessel can be located with high fidelity by using both the structural and angiographic contrasts. This technical advance lays a foundation for the absolute sO2 calculation in small vessels in human retina in order to assess local oxygen extraction and metabolism.

Non-invasive, near infrared angiography by high-speed digital holography allows for blood flow imaging and measurement of quantitative hemodynamic parameters in the human eye with an unrivaled temporal resolution. It reveals blood flow contrasts in the retinal and choroidal vasculature and enables vessel type identification. We report the first results of a clinical trial involving laser Doppler holographic monitoring of the eye fundus (IMA-MODE), designed to assess the pathophysiology of a number of ocular and general diseases appearing with age, such as occlusions, hypertension and diabetes, which may affect the ocular circulation, and provoke potentially blinding microvascular damage.

Through biomedical engineering partnerships, our group has been able to develop and translate optical coherence tomography (OCT), an imaging tool conventionally used in the “eye imaging suite”, for bedside use in unstable infants or for intraoperative guidance for the ophthalmic surgeon. In both arenas, image-guidance is transforming patient care. In the nursery, this has enabled assessment of retinal development, markers of injury and a determination of their relationship to brain function. In surgery, efficient application of an investigational high-speed swept-source OCT into our surgical armamentarium provides unique guidance in simple and complex cases, is improving our understanding of microsurgical pathologies, and is enabling new and improved surgical techniques.

Tunable laser sources with sweep-rates higher than 1MHz recently became commercially available. Today’s commercial ophthalmic OCT systems use sweep-rates in the 100-200kHz regime. These much faster laser sources can be used to either significantly reduce the imaging time or significantly increase the field of view (FOV). In this study we investigate the clinical value of OCT with MHz-rate swept source lasers. We implemented a versatile ophthalmic OCT system using a Frequency-Domain-Mode-Locked (FDML) laser with a sweep-rate of 1.7MHz, to address a variety of ophthalmic OCT imaging applications, exhibiting large imaging depth for wide field retinal OCT and OCT angiography (OCTA) with a field of view of up to 90 degrees, as well as for anterior segment imaging, and microscopic OCTA of the choriocapillaris with repetition rates of more than 1kHz.

We use the prototype swept-source (SS) OCT at 1060 nm to image human Schlemm’s canal (SC) and perform full reconstruction in the en face plane. Compared with spectral-domain (SD) OCT systems at 800 nm, the SS OCT system at 1060 nm offers deeper signal penetration, and has no sensitivity roll-off effect to allow for better localization and delineation of SC. One volumetric scan was taken from each of the eight cardinal positions to cover the entire SC circumferentially around the limbus. The en face slices were taken from each volume at the SC region, and were stitched together to generate an en face representation of the 360 deg SC.

Transparency of ocular structures affects contrast in the retinal image and has an impact on visual quality. Vitreous constitutes the largest volumetric component of the human eye, thus it contributes to the intraocular scattering. The vitreous can contain subtle opacifications causing an increase in scattering and a reduction in vision. We report three-dimensional enhanced depth imaging of the anterior vitreous with SS-OCT. We show visualization of anterior vitreous opacities (floaters). We also demonstrate the quantification of vitreal opacities with respect to the age of the subjects.

Bruch’s membrane (BM) and retinal pigment epithelium (RPE) changes, thought to be the earliest signs of impending atrophy in dry age-related macular degeneration (AMD), are subtle and challenging to detect non-invasively. Here, we demonstrate imaging and quantification of BM and RPE thickness in the in vivo human eye using achromatized visible light optical coherence tomography (OCT) in normal human subjects. Our consistent visualization of BM and agreement of thickness values with ranges from prior histological studies is attributed to the increased axial resolution, and potentially also to the enhanced contrast of BM in the visible light range.

In this paper, an automatic strategy for measuring the thickness of the nerve fiber layer around the optic nerve head is proposed. The strategy presented uses two independent 2D U-nets that each perform a segmentation task. One network learns to segment the vitreous body in standard Cartesian image domain and the second learns to segment a disc around a point of interest in a polar image domain. The output from the neural networks are then combined to find the thickness of the waist of the nerve fiber layer as a function of the angle around the center of the optic nerve head in the frontal plane. The two networks are trained with a combined data set that has been captured on two separate OCT systems (spectral domain Topcon OCT 2000 and swept source Topcon OCT Triton) which have been annotated with a semi-automatic algorithm by up to 3 annotators. Initial results show that the automatic algorithm produces results that are comparable to the results from the semi-automatic algorithm used for reference, in a fraction of the time, independent of the annotator. The automatic algorithm has the potential to replace the semi-automatic algorithm and opens the possibility for clinical routine estimation of the nerve fiber layer. This would in turn allow the detection of loss of nerve fiber layer earlier than before which is anticipated to be important for detection of glaucoma.

Ocular morphological deformations, such as scleral flattening due to high intracranial pressure, result in both global and more difficult to detect local changes. Local curvature analysis of the retinal profile could help in the visualization of retinal morphological deformations. Using a wide-field, whole eye OCT system, we demonstrate quantitative local, meridional retinal curvature maps across a 12 mm field-of-view. Because our method provides absolute (instead of relative) quantitative maps of local curvature, comparisons can be made directly between individuals. Additionally, retinal topography could help clinicians detect subtle, local retinal deformations earlier in patients and track changes over time.

Ophthalmic surgery is typically performed through an en-face only surgical microscope that provides limited depth information. This work introduces a high speed (400 kHz) microscope integrated optical coherence tomography (MIOCT) system which provides real time volumetric “4D” visualization via a heads-up stereoscopic display. The MIOCT system provides sub retinal visualization of tools and enables surgeons to perform delicate manipulation of retinal structures during mock surgical procedures. Following these mock surgical procedures in porcine eyes, this system will be readily translated into human ophthalmic microsurgery.

We investigate the influence of the OCT system resolution on high-quality en face corneal endothelial cell images in vivo, to allow for quantitative analysis of cell density. We vary the lateral resolution of the ultrahigh-resolution (UHR) OCT system (centered at 850 nm) by using different objectives, and the axial resolution by windowing the source spectrum. We are able to obtain a high-quality en face corneal endothelial cell map in vivo using UHR OCT for the first time. Quantitative analysis result of cell density from in vivo en face corneal endothelial cell map agrees with previously reported data.

Currently, clinical in-vivo imaging of the human limbus cellular structure is only possible with in-vivo confocal microscopy (IVCM). However, IVCM requires physical contact with the imaged object, and may cause incidental tearing or inflammation of the limbal tissue. We present a line-field, spectral-domain OCT system (LF-SD-OCT) that can generate volumetric, cellular resolution images of biological tissue in-vivo and without contact. The system provides 1.7 µm axial and 2.2 × 3.1 µm lateral resolution in tissue and 2.5 kHz frame rate. The quality of healthy human limbus images acquired with LF-SD-OCT is comparable to that of IVCM.

Quantitative features of individual ganglion cells (GCs) are potential paradigm changing biomarkers for improved diagnosis and treatment monitoring of GC loss in neurodegenerative diseases like glaucoma and Alzheimer’s disease. The recent incorporation of adaptive optics (AO) with extremely fast and high-resolution optical coherence tomography (OCT) allows visualization of GC layer (GCL) somas in volumetric scans of the living human eye. The current standard approach for quantification – manual marking of AO-OCT volumes – is subjective, time consuming, and not practical for large scale studies. Thus, there is a need to develop an automatic technique for rapid, high throughput, and objective quantification of GC morphological properties. In this work, we present the first fully automatic method for counting and measuring GCL soma diameter in AO-OCT volumes. Aside from novelty in application, our proposed deep learningbased algorithm is novel with respect to network architecture. Also, previous deep learning OCT segmentation algorithms used pixel-level annotation masks for supervised learning. Instead in this work, we use weakly supervised training, which requires significantly less human input in curating the training set for the deep learning algorithm, as our training data is only associated with coarse-grained labels. Our automatic method achieved a high level of accuracy in counting GCL somas, which was on par with human performance yet orders of magnitude faster. Moreover, our automatic method’s measure of soma diameters was in line with previous histological and in vivo semi-automatic measurement studies. These results suggest that our algorithm may eventually replace the costly and time-consuming manual marking process in future studies.

An adaptive optics optical coherence tomography (AO-OCT) system with a 3-sided pyramid wavefront sensor (P-WFS) is presented. Compared to the Shack-Hartmann WFS, the P-WFS promises better sensitivity in low-light scenarios and greater flexibility in pupil sampling. Key feature of the presented set-up is that part of the imaging light is used to illuminate the WFS. The double pass enables closed loop correction without beam modulation and speckle patterns in the sensor read-out are averaged out during scanning. The developed instrument is demonstrated with retinal images obtained in-vivo, where the cone mosaic is clearly visualized at ~4° eccentricity from the fovea.

To achieve 3D high-cellular resolution, a great effort, in the past years, was made to develop Adaptive Optics (AO)-OCT systems. However, such systems require quite complex, expensive and cumbersome hardware, making clinical transfer challenging. Moreover, conventional AO correction is limited to the retina isoplanatic patch (about 1deg), reducing the useful field of view (FOV). Recently, we showed the potential of Time Domain Full-Field-OCT (FF-OCT) to achieve 2deg FOV high-cellular-resolution in-vivo retinal imaging without using AO. Nevertheless, the technique was still facing some challenges in providing consistent and reproducible images (mainly due to axial eye motion), and it presented a reduced signal and FOV. Here, we present the new generation of FF-OCT system with axial eye motion tracking. We show new results and methods to minimize dispersion and coherence gate curvature enabling achieving 4deg (potentially 8deg) FOV high-cellular-resolution retinal images.

The biomechanical properties of the crystalline lens play a crucial role in its visual function. Assessing biomechanical properties of the lens may help with early disease detection and robust assessment of therapeutic interventions. However, measuring the biomechanical properties of the lens is a challenge due to its location inside the eye-globe. In this study, we demonstrate the combination of optical coherence elastography (OCE) and Brillouin microscopy to evaluate the stiffness of porcine lenses ex vivo (N=6). Brillouin microscopy can map the Brillouin-derived longitudinal modulus of the whole lens, but imaging times are lengthy. OCE can provide quantitative measurements of viscoelasticity rapidly, but the limited scattering of the lens limits its in-depth measurements. By combining these two techniques, we show a strong correlation between the Brillouin modulus and OCE-measured Young’s modulus in the lens, enabling depth-wise mapping of the Young’s modulus. The correlation coefficient between the two measurements was R=0.89. Using this correlation, the elasticity of the anterior lens was 2.72±0.89 kPa, and the mean Young’s modulus of the nucleus was 12.92±2.75 kPa. Similarly, the elasticity of the posterior lens was 3.80±1.25 kPa. While both techniques can evaluate the stiffness of the biological tissues separately, our work demonstrates that combining these techniques could enable mapping of the Young’s modulus completely noninvasively in non-scattering tissues such as the crystalline lens.

Tissue morpho-mechanics is gaining increasing in various fields, because it targets the relationship between morphological features and mechanical properties in biological tissues, which plays an important role in various fields including biology, medicine, pathology, tissue engineering, and regenerative medicine. The intimate connection between morphological, biochemical and mechanical properties in biological tissues requires a multimodal correlative approach for their exhaustive investigation. In this study, we used Second-Harmonic Generation in combination with Brillouin and Raman micro-spectroscopy in order to correlate collagen morphology at the ultrastructural level with its biomechanical and biochemical features. In particular, by imaging human corneal tissue samples with our multimodal approach, we demonstrated that the peculiar mechanical properties of corneal lamellae in the anterior portion of the corneal stroma are due to a different supramolecular organization, rather than to a different biochemical composition. This result opens new insights in the study and interpretation of corneal biomechanics thanks to the possibility of non-invasively correlating lamellar morphology with visco-elastic properties. The proposed method opens the way to the non-invasive assessment of corneal morpho-mechanics and holds the potential to be used for both diagnostic and follow-up purposes in pathologies that affect corneal biomechanics.

Air-puff induced corneal deformation imaging reveals information highlighting normal and pathological corneal response to a non-contact mechanical excitation. Here, we present a novel customized swept-source optical coherence tomography system coupled with a collinear air-puff excitation. We acquired unobstructed dynamic corneal deformation on multiple meridians with two custom scan patterns over a field of view of up to 15 mm x 15 mm and selected puff profiles at unprecedented scan rates, both ex vivo and in vivo. We show that our system can detect corneal deformation profiles and deformation asymmetries that are useful for corneal biomechanics diagnostics and pathology screening.

Nearly all benchtop studies of corneal biomechanics have relied on protocols which stiffen the cornea, such as riboflavin-UV crosslinking, as a way of providing contrast and validation of biomechanical measurements. However, there are strong clinical motivations to detect softening of the cornea. In this work, we present the evidence that phase-decorrelation OCT (PhD-OCT) is able to detect a small degree of corneal softening due to enzymatic digestion. This benchtop study supports the idea that PhD-OCT may detect keratoconus and early ectasia clinically.

Adaptive optics scanning laser ophthalmoscopy (AOSLO) has advanced the study of retinal structure and function by enabling in vivo imaging of individual photoreceptors. Most implementations of AOSLOs are large, complex tabletop systems, thereby preventing high quality photoreceptor imaging of patients who are unable to sit upright and/or fixate for an imaging session. We have previously addressed this limitation in the clinical translation of AOSLO by developing the first confocal handheld AOSLO (HAOSLO) capable of cone photoreceptor visualization in adults and infants. However, confocal AOSLO images suffer from imaging artifacts and the inability to detect remnant cone structure, leading to ambiguous or potentially misleading results. Recently, it has been shown that non-confocal split-detection (SD) AOSLO images, created by the collection of multiply backscattered light, enables more reliable studies of retinal photoreceptors by providing images of the cone inner segment. In this paper, we detail the extension of our HAOSLO probe to enable multi-channel light collection resulting in the first ever multimodal handheld AOSLO (M-HAOSLO). Imaging sessions were conducted on two dilated, healthy human adult volunteers, and M-HAOSLO images taken in handheld operation mode reveal the cone photoreceptor mosaic. Aside from being the first miniaturized and portable implementation of a SD AOSLO system, M-HAOSLO relies on sensorless optimization of the wavefront to correct aberrations. Thus, we also show the first ever SD images collected after correction of the eye’s estimated wavefront.

We developed a high-speed adaptive optics, line-scan spectral domain OCT and used it to characterize stimulus-induced optical path length changes in cones with high spatiotemporal resolution. We find that individual cone outer segments exhibit a biphasic light-induced response—a rapid axial shrinkage followed by a gradual increase in optical path length, both increasing in magnitude with the stimulus intensity. AO line-scan OCT thus offers high-speed volume acquisitions, high phase stability, sub-ms temporal resolution and cellular-scale spatial resolution, that together enable imaging retinal structure and function in health and disease.

In vivo functional imaging of human photoreceptors is an emerging field, with compelling potential applications in basic science, translational research, and clinical management of ophthalmic disease. Measurement of light-evoked changes in the cone photoreceptors has been successfully demonstrated using adaptive optics (AO) coherent flood illumination (CFI), AO scanning light ophthalmoscopy (SLO), AO optical coherence tomography (OCT), and full-field OCT with digital aberration correction (DAC). While the optical and computational principles of these systems differ greatly, and while these differences manifest in the resulting measurements, we believe that the approaches are all sensitive to light-evoked swelling of the cells. We describe a combined OCT-SLO with AO designed to measure this light-evoked swelling. In addition to OCT measurement of cone responses, we report their simultaneous OCT-SLO measurement as well as OCT measurement of rod photoreceptor function, neither of which, to our knowledge, have been reported before.

Retinitis Pigmentosa (RP), the most common group of inherited retinal degenerative diseases, is characterized by progressive loss of peripheral vision that surrounds an island of healthy central vision and a transition zone of reduced vision. The most debilitating phase of the disease is cone photoreceptor death whose biological mechanisms remain unknown. Traditional clinical methods such as perimetry and electroretinography are gold standards for diagnosing and monitoring RP and indirectly assessing cone function. Both methods, however, lack the spatial resolution and sensitivity to assess disease progression at the level of individual photoreceptor cells, where it begins. To address this need, we developed an imaging method based on phase-sensitive adaptive optics optical coherence tomography (PS-AO-OCT) that characterizes cone dysfunction in RP subjects by stimulating cone cells with flashes of light and measuring their resulting nanometer-scale changes in optical path length. We introduce new biomarkers to quantify cone dysfunction. We find cone function decreases with increasing RP severity and even in the healthy central area where cone structure appears normal, cones respond differently than cones in the healthy controls.

Microglia are central nervous system macrophages and the first responders to neural injury. Herein we characterize their distribution and motility in human eyes using a multimodal AO system. In healthy eyes, microglia are absent in the central macula up to ~5º eccentricity but their density increases monotonically at higher eccentricities. Microglia density decreases linearly with age. ILM microglia are relatively immobile for durations up to two weeks but their processes re-orient over timescales as short as minutes. The density, motility, and reactive state of microglia may serve as an ocular disease biomarker for early detection and progression monitoring.

Over last few years Optical Coherence Tomography has allowed precise observations of the light-evoked opthophysiological response of the inner and outer retina. In mouse rods this manifests itself as outer segments elongation that is driven by phototransduction. Given the current model describing these observations as osmotically driven water movements we have decided to alter retinal water homeostasis to see if, and how the measured signals will be affected. Our observations of light-evoked retina responses during mild, diffuse central retinal edema confirmed that water plays a key role in this process.

Recent Alzheimer’s disease (AD) studies have focused on retinal analysis, as the retina is the only part of the central nervous system which can be imaged non-invasively by optical methods. The appearance of the retina of the APP/PS1 mouse model of AD was therefore evaluated by a custom-built multi-contrast optical coherence tomography (OCT) system. By visualizing retinal changes over a range of ages in intensity-, polarization- and motion-based contrast modes simultaneously, observations in the retina were documented and compared to both retinal whole-mounts and conventional histological sections. The data was also correlated to the cortical amyloid beta plaque load.

A combined scattering and autofluorescence (AF) measurement from the retinal pigment epithelium (RPE) was investigated to study the changes in the AF granules in an Abca4-/- mouse (murine model of Stargardt disease). The directional scattering was measured with an optical coherence tomography (OCT), whereas the multi-color fundus AF spectra was measured with a spectrometer-integrated scanning laser ophthalmoscope. Increased scattering from the RPE in Abca4-/- mouse relative to control was well correlated with the elevated AF spectra and indicated the changes in RPE fluorophores. Ex vivo studies based on scanning confocal and electron microscopy were performed to validate the in vivo findings.

Mice are commonly used to model human retinal disease because of similarities in ocular anatomy and function and availability of transgenic phenotypes. Recently, targeted delivery of photo-lesions has been demonstrated using fundus imaging and a combined optical coherence tomography (OCT) and scanning laser ophthalmoscopy (SLO) system; however, these systems precluded simultaneous multimodality imaging. Here, we present multimodality OCT+SLO imaging combined with a photocoagulation laser and custom software that allows for targeted delivery of focal retinal laser injury under real-time en face OCT guidance. Longitudinal imaging results show lesions with varying degrees of retinal damage during an injury titration study.

Diabetes is a chronic metabolic disease characterized by elevated levels of blood glucose. Over time, it can lead to serious damages in the body. In the eyes, diabetic retinopathy (DR), the most common microvascular complication of diabetes, is a major cause of blindness. OCT can be used to provide high resolution images of the damages in the retina and follow their evolution over time. It is however still unclear which of the vascular or neurologic changes happen first in the development of the disease. In this work, we investigate the birefringence of the retinal nerve fiber layer (RNFL) of diabetic patients (with different stage of DR or no DR) and compare these results to healthy subject’s data. We use a PS-OCT system with an integrated retinal tracker for imaging (center wavelength of 860 nm, A-scan rate of 70 kHz). For each eye imaged, a raster scan centered on the optic nerve head (ONH) and a circular scan around the ONH (radius: 1.5mm) are taken. Considering only areas with a RNFL thickness >100 μm, birefringence values are calculated from an averaged circular tomogram for each eye. We observe a statistically significant reduced birefringence of the RNFL in the diabetic patients compared to the healthy volunteers.

We used multi-contrast OCT (MC-OCT), which is capable of the simultaneous measurement of OCT angiography, degree of polarization uniformity and intensity OCT, to evaluate retinal pigment epithelium (RPE) changes. MC-OCT system was operated at an axial scan speed of 100,000 A-scans/s, using a swept-source laser at a central wavelength of 1,048 nm. From the dataset of MC-OCT, a pixel-wise segmentation method for RPE-melanin was developed and used to create RPE-melanin-specific contrast images to evaluate RPE-melanin changes. The RPE-melanin cross-sectional images were generated to evaluate the depth-resolved distribution of RPE-melanin. RPE-melanin thickness maps were created by counting the number of pixels with RPE-melanin at each A-line in the 3D dataset. An RPE-melanin thickness map represents the en face distribution of the thickness of RPE-melanin. We evaluated 37 eyes with age-related macular degeneration (AMD) with serous retinal pigment epithelium detachment, and 24 eyes with chronic Vogt-Koyanagi- Harada (VKH) disease. In these cases, RPE-melanin thickness maps showed similarities to the near infrared autofluorescence (NIR-AF; excitation 780 nm) images. In the eyes with AMD, focal RPE damages could be readily detected with RPE-melanin thickness map. RPE-melanin cross-sectional images were more sensitive for the damage at RPE-Bruch’s membrane band than intensity OCT images. In the eyes with VKH disease, RPE-melanin-specific contrast images clearly showed focal RPE-melanin accumulation at granular hyper NIR-AF lesions. In conclusion, this study demonstrated the clinical usefulness of RPE-melanin specific contrast OCT imaging for evaluating RPE changes in retinal diseases.

Polarization sensitive optical coherence tomography (PS-OCT) has been used to visualize the orientation of the nerves in the retinal nerve fiber layer (RNFL) and to visualize depolarization in retinas of healthy volunteers and age-related macular degeneration (AMD) patients. Optic axis orientation images clearly visualize the nerve fibers leaving the optic nerve head (ONH) in all radial directions in healthy volunteers. Depolarization images show depolarization of the RPE and for some cases, highlight another depolarizing layer at the boundary of the choroid and sclera.

Fundus imaging is widely used for the diagnosis of retinal diseases. Major ophthalmic diseases like glaucoma, diabetic retinopathy (DR), age-related macular degeneration (AMD) are diagnosed by examining retinal fundus images. Therefore, the efficient and reliable diagnosis largely depends upon the resolution of the images. In different diseased conditions, different pathologies and landmarks (haemorrhages, microaneurysms, exudates, blood vessels, optic disc and optic cup, fovea) of the retina get affected. In clinical situations it is often not possible to obtain good high-resolution images. Here, the techniques of super-resolution can be applied. The objective of super-resolution is to obtain a high-resolution image from a low-resolution input image. In this paper, we present results of the application of enhanced deep residual networks for single image super-resolution (EDSR) on retinal fundus images. This network is based on the SRResNet architecture involving skip connections. Using the public RIGA dataset, which consists of glaucoma and normal fundus images, we have trained the model using 2x, 4x and 8x scaling with three different optimizers each (namely ADAM, Stochastic Gradient Descent and RMSprop) to determine which optimizer is best for the different scales. We have also provided results obtained by varying the residual blocks in the network.

Dry eye is one of the most reported eye health conditions and it is characterized by dryness, decreased tear production, or increased tear film evaporation. Middle-aged and elderly people are most commonly affected because of the high prevalence of contact lens usage, systemic drug effects, autoimmune diseases, and refractive surgeries. Corneal topography images have been recently used for noninvasive assessment, based on the Placido rings pattern. The rings in normal eyes are smooth and have no distortion, whereas they are distorted in affected eyes. We developed a method of analysis that process the corneal topography image to determine the Tear Break-up Time (TBUT), using the Tear Film Surface Quality (TFSQ) measurement. To avoid distortions not caused by the tear film break-up, the method dynamically removes eyelashes shadows from the image processing area. The results show that the proposed analysis is able to determine the TBUT based on the graphical analysis, and it can be used to help eye care specialists to diagnose dry eye disease.

There is a strong association between exposure to blue light (from 400 nm to 500 nm), especially high-energy blue light (from 400 nm to 450 nm), and damage to ocular structures. And the use of blue filters can prevent the frequency of eye damage. This study aims to develop and build a prototype for measuring protection against blue light in sunglasses and to determine whether this protection is adequate, adopting as criterion its blue-light transmittance be less than 1.2 times its luminous transmittance. To obtain the proper spectral weighting functions, the measurement will consist of linear combinations of the responses of a TCS3472 photosensor illuminated by a high-brightness white LED. The agreement between the measurements made with the proposed device and a gold standard will be analyzed by the Bland-Altman method. The developed device will be available to public to measure their own sunglasses. Moreover, this device has potential for use in the trade to add value in sales of sunglasses and in the industry for quick conference of sunglasses produced.

To quantitatively describe and evaluate a new image processing technique for estimating the Foveal Avascular Zone (FAZ) in subjects with Diabetic Retinopathy and myopes. From a total of 328 images obtained from Diabetic Retinopathy (113), myopes (120) and normal (93), the FAZ dimensions were quantified using a new image processing algorithm. These parameters were also determined manually and by the OCT manufacturer’s inbuilt algorithm. In the new technique, the images were first pre-processed by using a DOG filter iteratively before being complemented followed by a Prewitt edge detection and repeated image dilation at angles of 00, 450 and 900. Image closure was then applied followed by noise and small object removal which resulted in the segmented boundary. For deeper insight into shape change, in addition to the diameter of the FAZ other parameters such as the area, diameter, major axis, minor axis, orientation, perimeter vessel avascular density (VAD), Vessel diameter Index (VDI), etc. were obtained. The circularity index was calculated using the FAZ area and perimeter parameters. The mean FAZ diameter (mm) by the new automated technique, manual-segmentation (ground truth), and inbuilt instrument algorithm were 0.67 ± 0.87, 0.67 ± 0.72 and 0.61 ± 0.14. The mean of FAZ area (mm2) was 0.36 ± 0.10, 0.33 ± 0.09 and 0.43 ± 0.14 in normal, myopia and diabetic subjects respectively. The new technique shows considerable improvement in accuracy (mean ± SD) when compared to the inbuilt system segmentation and the ground truth (manual marking by an expert clinician). The study results show that the FAZ area in Diabetic Retinopathy is significantly different (p=0.003) when compared to myopic eyes (p=0.016) and normals.

Early detection of diabetic retinopathy (DR) is an essential step to prevent vision losses. This study is the first effort to explore convolutional neural networks (CNNs) for transfer-learning based optical coherence tomography angiography (OCTA) detection and classification of DR. We employed transfer-learning using a pre-trained CNN, VGG16, based on the ImageNet dataset for classification of OCTA images. To prevent overfitting, data augmentation, e.g. rotations, flips, and zooming, and 5-fold cross-validation were implemented. A dataset comprising of 131 OCTA images from 20 control, 17 diabetic patients without DR (NoDR), and 60 nonproliferative DR (NPDR) patients were used for preliminary validation. Best classification performance was achieved with fine-tuning nine layers of the sixteen-layer CNN model.

Most sunglasses standards require ultraviolet protection from 280 nm – 380 nm to ensure the limits for effective spectrally weighted radiant exposure. They are a mirror of ISO 12312-1 and do not consider UV-A upper limit as 400 nm. Some standards have extended the UV-A limit, however, none of them considers the World Health Organization safe limits for unweighted radiant exposure: ultraviolet radiant exposure in the spectral region 180 nm – 400 nm incident upon the unprotected eye(s) should not exceed 30 J/m² effective spectrally weighted (based on the actinic spectra), and the total (unweighted) ultraviolet radiant exposure in the spectral region 315 nm to 400 nm should not exceed 10 kJ/m2. Calculations of these limits were performed for 27 Brazilian state capitals, and they led to a change in the upper UV-A limit to 400 nm on the 2013 review of the Brazilian standard NBR 15111:2013. Moreover, because the sunlight irradiance in Brazil is quite high, integration over the 280 nm – 400 nm range yields an ultraviolet radiant exposure (spectrally weighted) that is an average of 49% greater than that for the 280 nm – 380 nm range. Additionally, the unweighted ultraviolet radiant exposures are over the limit. Ultraviolet radiant exposure is latitude dependent. Brazil exceeds the limit of unweighted radiant exposure on average 40% – 50%. Europe and the US have a similar scenario and exceed around 40%. Furthermore, despite the blue light hazards mentioned in the literature, there are no limits whatsoever established on any sunglasses standards, and our calculations show that not only they exceed the safety limits, but they should also be included on the standards.

The Foveal Avascular Zone (FAZ) is of clinical importance since the vascular arrangement around the fovea changes with disease and refractive state of the eye. Therefore, it is important to segment and quantify the FAZs accurately. Here we provide a new methodology for this measurement. Eighty normal fundus images of dimensions 420x420 pixels corresponding to 6mm x 6mm were used in this study. Each fundus image was manually segmented by a clinical expert (ground truth), the new methodology and an existing technique provided by the image acquisition device (Cirrus 5000 Carl Zeiss Meditec Inc., Dublin, CA). The images were first processed by a Difference of Gaussian (DoG) filter iteratively 25 times after being complemented. This is followed by a Prewitt edge detection and repeated image dilation at angles of 0,45 and 90 degrees. Image closure was then applied followed by noise and small object removal which resulted in the segmented boundary. For deeper insight into shape change, besides the diameter of the FAZ other parameters - eccentricity, perimeter, major axis, minor axis, incircle, circumcircle, Fmin, Fmax, tortuosity, vessel diameter index and vessel avascular density - were calculated. The mean diameter by manual segmentation was 673.04 ± 86.92 μm compared to 688.42 ± 72.18 μm by our technique. The corresponding value generated by the instrument was 623.60 ± 121.50 μm. This technique shows considerable improvement in accuracy (the mean value as well as the standard deviation) when compared to system segmentation and the ground truth. These aspects will be discussed in the paper.

In this study, a diversity of the human lacrimal canaliculus (LC) shape was shown using dynamic optical coherence tomography (D-OCT) method. D-OCT is a method of clear imaging of a fluid using a contrast agent. LC plays an important role in tear drainage system and it is expected to assess LC in detail without pain of subjects. Non-invasive and non-contact OCT imaging of LC of 6 eyes of 3 subjects was performed. The LC images were cut out by applying the D-OCT method and three-dimensionally reconstructed to evaluate the characteristics of LC. Although no significant difference was found between the left and right LCs of the same subject, it was shown that individual differences were remarkable.

We present a new wave front sensing technique based on detecting the propagating light waves. This allows the user to acquire millions of data points within the pupil of the human eye; a resolution several orders of magnitude higher than current industry standard ophthalmic devices. The first instrument was built and tested using standard calibration surfaces in addition to using an artificial eye. The paper then presents the first characterization of the optics of a real human eye measured using the newly developed high-resolution wave front phase sensing technique showing the complexity of the human eye’s ocular optics.

Wavefront metrics such as root mean squared error provide excellent descriptions of optical quality but the connection to visual performance is not directly interpretable from the wavefront alone. Converting to visual acuity (VA) would provide a more accessible assessment of the effect of ocular wavefront. In this study, multiple measurements of wavefront and pupil diameter were acquired at 1 cd/m² using the iDesign 1.3 (Johnson and Johnson Surgical Vision) in 552 eyes of 293 subjects. Uncorrected (UC) and best corrected (BC) VA were measured at 4m for each eye at 100 cd/m² . VA was estimated using a neural contrast sensitivity function (NCSF) weighted modulation transfer function (MTF). To estimate BCVA, sphere and astigmatism terms were nulled. For each subject, classification was performed using random forest considering estimated VA, age, measured pupil diameter, manifest refraction spherical equivalent, manifest refraction cylinder and gender. Across all measurements, predicted differed from measured UCVA by -0.11 and BCVA by +1.6 (-10logMAR). The NCSF-based model predicts population mean VA to within approximately 2-lines but cannot predict an individual’s VA well, supporting the notion that other factors need to be considered to obtain more accurate estimates of VA. Classification with the random forest approach improved the accuracy of estimates of an individual subject’s VA; approximately 95% of the estimates match the measured VA.

Fluorescein videoangiography is often used to visualize diabetic retinopathy (DR), a degenerative disease characterized by degradation of retinal blood vessels. However, more sensitive and quantitative measures of increased vessel permeability in the asymptomatic phase of DR (non-proliferative DR) could help identify patients who would benefit from therapeutic intervention to avoid vision loss. Here, a modified “adiabatic approximation to the tissue-homogeneity model” was shown in simulations and rat experiments to estimate extraction fraction within a 20%-error for physiological ranges of vascular permeability expected in nonproliferative DR, even in low-dynamic-range (8-bit) fluorescein imaging systems that are standard in many institutes.

Diabetic retinopathy (DR) is a major ocular manifestation of diabetes. DR can cause irreversible damage to the retina if not intervened timely. Therefore, early detection and reliable classification are essential for effective management of DR. As DR progresses into the proliferative stage (PDR), manifestation of localized neovascularization and complex capillary meshes are observed in the retina. These vascular complex structures can be quantified as biomarkers of transition of DR from no-proliferative to proliferative stage (NPDR). This study investigates four optical coherence tomography angiography (OCTA) features, i.e. vessel complexity index (VCI), fractal dimension (FD), four-point crossover (FCO), and blood vessel tortuosity (BVT), to quantify vascular complexity to distinguish NPDR from PDR eyes. OCTA images from 20 control, 60 NPDR and 56 PDR patients were analyzed. The univariate analysis showed that, with the progression of DR, all four complexity features increased with statistical significance (ANOVA, P < 0.05). A posthoc study showed that, only VCI and BVT were able to distinguish between NPDR and PDR. A multivariate logistic regression identified VCI and BVT as the most significant feature combination for NPDR vs PDR classification.

Photoacoustic microscopy (PAM) is a non-invasive and hybrid optical imaging technique that has a potential to visualize chorioretinal vasculature in vivo. The capability of PAM can be extended to better visualize the dynamic changes of the vasculature network in the retina when it is combined with another imaging modality such as fluorescence microscopy or OCT. In this study, an integrated PAM and OCT has been developed to identify the local tissue damage during laserinduced photocoagulation on major choroidal vessels. Choroidal lesion was induced using a high power green light laser at 532 nm with millisecond pulse duration in eight New Zealand rabbits. Each rabbit eye was irradiated for 0.5 s at a laser power of 750 mW and spot size of 100 μm. Six laser burn positions were created on each eye. At each laser burn, twenty shots of the laser were applied. Multimodal PAM, OCT, fundus, and FA were used to monitor thermal lesion at different time points (days 0, 1, 3, 5, 7, 14, 21, and 28) after photocoagulation. All thermal lesions were clearly identified with high resolution using PAM. In addition, the PAM images exhibited dynamic changes of density and morphology of choroidal vasculature. The OCT images provided visualization of the cross-sectional structure of retinal tissues and the location of thermal lesion. Multimodal PAM and OCT can provide a feasible tool for evaluation and monitoring of damaged tissues and the microvasculature of the retina.

Corneal pathologies are leading causes of blindness and represent a world health problem according to the world health organization. Early detection of corneal diseases is necessary to prevent blindness. In this paper, we use transfer learning with pretrained deep learning networks to diagnose three common corneal diseases, namely, dry eye, Fuchs' endothelial dystrophy, and keratoconus as well as healthy eyes using only optical coherence tomography (OCT) images. Corneal OCT scans were obtained from 413 eyes of 269 patients and used to train, validate, and test the networks. All networks achieved all-category accuracy values > 99%, categorical area under curve values > 0:99, categorical specificity values > 99%, and categorical sensitivity values > 99% on the training, validation, and testing, respectively. The work in this paper has clinical significance and can potentially be applied in clinical practice to potentially solve a significant world health problem.